An Overview of Feature-Based Shadow Detection Methods

An Overview of Feature-Based Shadow Detection Methods

Suhaib Musleh (Kuwait University, Kuwait), Muhammad Sarfraz (Kuwait University, Kuwait) and Hazem Raafat (Kuwait University, Kuwait)
DOI: 10.4018/978-1-7998-4444-0.ch007
OnDemand PDF Download:
Available
$37.50
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

Shadows occur very frequently in digital images while considering them for various important applications. Shadow is considered as a source of noise and can cause false image colors, loss of information, and false image segmentation. Thus, it is required to detect and remove shadows from images. This chapter addresses the problem of shadow detection in high-resolution aerial images. It presents the required main concepts to introduce for the subject. These concepts are the main knowledge units that provide for the reader a better understanding of the subject of shadow detection and furthering the research. Additionally, an overview of various shadow detection methods is provided together with a detailed comparative study. The results of these methods are also discussed extensively by investigating their main features used in the process to detect the shadows accurately.
Chapter Preview
Top

Background

Various authors have worked in the area of shadow detection. For the methods that depend on image pixels intensity values and color information, Polidorio et al., in (Polidorio, Flores, Imai, Tommaselli, & Franco, 2003), exploited two properties of shadows, high blue/violet wavelength and low brightness to detect shadow in images. In the beginning, the RGB image was converted to HSI model using the appropriate equations. Then a partition process was applied on the saturation and intensity components to detect the shadows.

Complete Chapter List

Search this Book:
Reset